Working plotly integration

[metahertz]

May be of interest, may not.

I wanted something that would allow me to easily monitor link quality over time; this is pretty easy for latency etc but nothing gave me an automated way of graphing speedtests.

The changes simply add a --plotly option which, if used, will spit out the upload, download and latency measurements from the results dict to the online graphing service, subsequent runs will append to the existing graph, with the timestamps for the X axis also being taken from the results dict.

Example graph output (mine) is here: https://plot.ly/~matjohn2/6.embed

Have also added a little Dockerfile and info in the readme as to how I use this,
but like I said, may be of little interest to the wider speedtest-cli community, could have created a separate wrapper using the API, but only just noticed the API when submitting the PR ;)

[Mraanderson]

That's nice. I've been using speedtester-cli-extras to save results to a CSV then using that CSV as a live tab in a Libre office Calc ods file. On a separate tab a graph is made up of the results from the external source. Pretty basic.

[sivel]

This is not going to be accepted in this current form. I would entertain a contrib script that does this, but not integrated in the main speedtest.py file. The contrib script would need to live in a contrib directory, and it would also not be included in pip installs. It wold have to be manually downloaded from github.

[metahertz]

Hey @sivel thanks for the feedback!

I realized as soon as i found your API that this was a bad PR.
Instead, i'll build a separate script, calling speedtest-cli via it's APIs, then I can distribute via docker images directly (which was my primary need!).

Thanks for the project though, awesome and saved me a ton of time!

[sivel]

Please keep us up to date on your progress. I'd like to create a wiki page or similar to link to projects of this type.